can anyone using Debian's Chromium reproduce this bug?

And a good one at that. My service is advertised as 60mbit/s download. Sustained 64 for 49 hours straight.

@jasonwryan that's why you should just downvote/vtc...

To reboot on a debian-based system, I typically do 'reboot' as root.. I know you can also do 'init 6' and it will reboot. Is there a different between the commands, i.e. one has a graceful shutdown while the other just stops everything?

Does anyone know what mistake I'm making here?

Never mind, apparently subprocess does not support this.

@FaheemMitha if you want to do that you should make sure you invoke by passing in shell=True, default is running without a shell. And I recommend you look at the plumbum package plumbum.readthedocs.org/en/latest if you do that kind of things from python

I need to know how KDE saves the information that the desktop should be shown as a directory-view

@salbeira - ask Q's on the main site, not in the chatroom!

@Anthon Ok, I will. Thanks.

@adywp would need to have you describe what you're trying to do better before attempting to guide you more. Also you Q sounds like it might be suited for here or on SO, tipping towards SO, given it may be more of a programming Q. However we need more to go on.

@ekaj init 6 and reboot result in the same actions. init 6 switches to runlevel 6, and on debian the last item to be run is K11reboot, which calls reboot. reboot is a symlink to halt and if you call it when not in runlevel 0 or 6, then halt invokes shutdown which for a reboot, switches to runlevel 6, accomplishing the same thing as init 6.

so in short, from a proper shutdown perspective there is no difference between reboot and init 6 although they take slightly different routes to the same endpoint.

Note however, that the concept of runlevels will be dissapearing with the upcoming debian systemd transition, so you might just want to get into the habit of using reboot rather than manipulating the runlevel directly.

Though I'd imaging they'll wire init 6 to still do the right thing for backward compatibility

yipee!!!!!!! 10k

@Ramesh congrats!!

@slm Thanks! Figuring out what new privileges I have now :)

seeing deleted A'ers

@Ramesh Congrats.

@FaheemMitha Thanks.

@Ramesh Yay!

@FaheemMitha Sorry for snapping at you yesterday, I was under a lot of stress and I lashed out at an innocent bystander.

I should be able to help today if you like.

@terdon No problem.

@terdon I think I've got things under control. Glad you're feeling better.

I heard back from EBL. They said the data is freely distributable, but there seemed to be a little ambiguity there. I can forward the email to you if you want.

sorry, that should be EBI.

@FaheemMitha As far as I know, their data are all in the public domain.

@terdon So, you think it is Ok to just include it with my code?

But explain to me what you're actually after here. Are you distributing a piece of software? Or do you just need to give the reviewers access to the sequences? I mean, will they need to be bundled with your software? Will the future users need the sequences?

@FaheemMitha I think so yes. IANAL of course.

@terdon Ok. I have some code, which needs example data sets. Those sequences are those example data sets. The code needs them to run. The code is online. You can see it at bitbucket.org/faheem/corrmodel

I sent you the paper corresponding to this earlier. You may not remember.

@FaheemMitha The one from a few months ago right?

Anyone who wants to check the results, not just the reviewers, will need those sequences/data.

@terdon Right.

I don't actually have any other data that the code runs with, so it's kind of indispensible.

@FaheemMitha That's not an issue. You just include the accessions as supplementary data. If those accessions are from EBI, a public database, I am 99% sure you can also include the sequences themselves.

@terdon You mean upload it to the journal as supplementary data?

I was going to just include it with the code, and send the journal a snapshot of the code.

Which is actually fairly useless, because there will be bug fixes and so forth, but better than nothing.

Maybe the data could/should be uploaded separately. I guess I'll ask the journal if/when I get that far.

@FaheemMitha Yes. Or in the downloadable archive that includes your code. I've seen that quite often for example. A particular piece of software comes with a directory called "examples" and they are sequences from a public database.

@terdon Right.

I'm a little cautious about including it in my repo, because once committed, it is not easily uncommitted. I'd have to rewrite the repo.

@terdon while we're on the subject, your roommater never got back to you about the paper? I remember you said he works in sort of the relevant area.

@FaheemMitha Yes, I passed him the mail but I don't know if he ever read it. I seem to recall him telling me that it was not actually that relevant to what he does.

@terdon Hmm, Ok. Well, reading research papers is quite a lot of effort.

@FaheemMitha I would include it. I find it very hard to believe that this lady has either any objections or any right to an opinion. I'm still pretty sure there was a misunderstanding and she does not get what you are suggesting.

@terdon Ok, thanks.

It seems EBI is covered by a blanket policy, so hopefully it will be Ok.

@FaheemMitha Yes, I think you will have no issues. Especially since the chances of anyone who is enough of a dick to care reading your paper are pretty slim. Lawyers don't tend to go for that kind of reading material :)

@terdon Very true.

I forgot if you told me. If you are leaving your current institute, do you know where you are going yet?

I think I'll forget to tell that lady professor when that paper is published.

@FaheemMitha if you have any reservations about distributing the data, you could always just include a script to fetch the data rather than include the data

@casey Yes, that's what I've been doing. But it is a little bit of a pain.

@FaheemMitha Not really, no. Kind of in limbo at the moment.

@FaheemMitha You should.

Writing a script, I mean.

Forget, I mean.

@terdon Right

@terdon I see. Is the job scene bad at the moment?

sometimes I hate data...

I gather you are still trying to finish up old projects. That can be rough.

@FaheemMitha Not as such, I need to figure out what I want to do and where I want to do it.

@terdon Ok.

@casey Details?

This one I can't figure out and it has bugged me for a long time. this code uses parallel hdf5 to write its output and writes them out at every history dump time (e.g. every 15 seconds of model time) and once its down writing a timestep, its done.

I actually thought bioinformatics was kind of hot/trendy, but I guess it depends on what you want to do.

but... if the model doesn't end cleaning and close the phdf5 libraries, the data is corrupted at all time steps

This is your code?

The hdf library is doing something critical at the end, but I have no idea what...

@FaheemMitha primarily no, and the issue is likely in the depths of the parallel hdf5 library. Normally this isn't a problem.

@casey Ok

but... I submitted a wrong wallclock time to the supercomputer and the model was killed at timestep 8920 out of 9000 and all that data is junk

I should add that it is parallel hdf5 via MPI. A lot of complexity there :/

@casey parallel code can be nasty to debug.

no time for that today. I'm just re-running the model with a higher wallclock limit

need the data ASAP :)

cause I'm presenting it in 10 days

ok

@terdon thanks. But now the OP unaccepted one of the answer and I should wait for some more time to get the privileges back :)

Is anyone here familiar with autotools/cmake?

@Ramesh Ah! I've been wondering about that, so you lose the privileges if you fall back under the threshold?

@terdon yeah. It was there when I went for lunch. Now it's gone :)

@Ramesh there you go.

@terdon yay!! Thank you..Now time to do some moderation stuffs :P

:)

@terdon J00545 seems to have disappeared. It was there in an earlier version of the data.

Any idea what happened to it?

That's the EBI accession number, I think.

@FaheemMitha It's been renamed to V00778

@FaheemMitha thing is sometimes you get the same sequence from different sources and different accessions and they're merged. Others, it turns out a sequence was an error and it gets deleted, then they change for various other reasons. It's hell./

@terdon Oh. How did you figure that out?

In this case, apparently dbfetch couldn't figure it out.

@FaheemMitha I went to the EBI website and pasted the accession into their search field => got the new one

@FaheemMitha I know, I spent months and months struggling with various names and ended up using a multipronged approach where I would use a webservice, then parse a local database I had built and then another webservice. I still can't get all names reliably.

@terdon Ah, obvious, but I didn't think of doing that.